GD&T

Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 46

Actual Mating Envelope

Calculating Additional Tolerance Permitted when a Position Tolerance is Modified at


MMC.
An additional tolerance, often called Bonus tolerance, is permitted when a position tolerance is
modified at maximum material condition (MMC) and the actual mating envelope of the feature has
departed from the MMC.
In these illustrations, the part is blac and the actual mating envelope is red.
For an internal feature such as a hole or slot:
For an external feature such as a boss or tab:
ote: !he Actual Mating "nvelope must be oriented relative to the specified #atums. $hen
an inspector merel% uses the si&e of a feature to calculate the bonus tolerance, out of spec
parts ma% be accepted.
If a hole, for instance, has the following si&e and geometric control, and the hole measures .'(). It
would be incorrect to use a bonus tolerance of .((* (.'() + .,--(MMC)) if the hole is not perfectl%
oriented to the #atums. If the hole is out of perpendicular to datum A b% .((), for instance, the
bonus that ma% be used is reduced b% that amount. !he bonus would be merel% .((. and the
allowable position tolerance / .(.0.
Angularit!
"se Two #atum $eferences for Angularit!%
$hen setting up a part for the measurement of angularit% on a sine bar, it is difficult to orient the part
before performing the inspection.
!his problem is overcome if a second datum is referenced in the angularit% callout

&asic #imensions
Another Basic Dimensions Tip
'ive Production the #imensions The! eed.
$hen 1#2! is used, the geometric tolerances appl% to the features + not the dimensions. !herefore when
BA3IC dimensions are used with geometric tolerances, several different dimensioning schemes ma% be used
without changing the meaning of the drawing. !he first drawing below is an example of baseline dimensioning.
!he other two illustrate chain dimensioning. 3ince the position tolerance is related to datums A, B and C, all
three drawings have the same meaning even though the dimensioning is different. !his would not be true if the
dimensions locating the holes were toleranced rather than BA3IC. Because the meaning of these drawings is
the same, the designer should consider the needs of those who will read the print when placing dimensions.

A(ME )*+.+* , -../ , 0earn to deal with &asic #imensions.
Another ASME Y14.41 - 2003 Tip
!here are two t%pes of dimensions+toleranced and basic. #imensions that are toleranced ma% be toleranced on the
field of the drawing, in a general note or a supplemental bloc of the drawing format. Basic dimensions are toleranced
indirectl% with geometric tolerances. !he geometric tolerance applies to the feature, not the basic dimension. $e have
4ule 5. in the A3M" 6.,.' standard which explains the meaning of a toleranced si&e dimension. 7or other than si&e,
toleranced dimensions can have multiple meanings. 7or that reason, the A(ME )*+.+*,-../ standard on digital
modeling recommends that direct tolerancing onl! be used to define the si1e of a feature 2/.*.*3 item4+56.
1eometric tolerancing is the preferred method. !he% do allow some direct tolerancing for other than si&e but in man%
cases recommend that the dimension origin s%mbol be used to clarif% the meaning.
8ere is a drawing made in accordance with the A3M" 6.,.,.+)((* standard. An% dimensions not shown ma% be
determined b% maing 9ueries of the digital model.

#igital product data definition, reduced dimension drawings and paperless part definition will continue to grow in
popularit%. !he sooner organi&ations embrace geometric tolerancing and use toleranced dimensions primaril% for si&e,
the sooner technolog% in this area will advance.

&oundaries
Thin7 &oundar!3 not &onus
!he t%pical approach to inspecting a feature such as the
n..((( hole is to measure the si&e, calculate a bonus and then
tr% to locate the axis of the hole to see if it is within the
allowable position tolerance. Another approach is to inspect
the si&e and mae measurements to see if the hole violates
the inner boundar% (virtual condition). If a hard gage were
produced to inspect the position of this hole, it would have a
pin n.-:( (;gage tolerance and wear allowance) located at the
basic location from the datums.
!his hard gage ma% be simulated using a height gage in an
open setup or CMM probe. B% adding and subtracting half the
virtual condition from the basic dimensions, the distance from
the datums that ma% not be violated ma% be calculated. !hese
distances ma% be easil% inspected.

Centerlines
#imension to Centerlines 8hen it Ma7es (ense
Man% past !ips have illustrated dimensioning from and to center lines. !here continues to be 9uestions about when it is
oa% to dimension with center lines and when it is not appropriate. #imensions ma% be to and from center lines on a
drawing if the dimensions are basic and the center line is representing the<
axis, center point or centerplane of a feature of si&e
center of a pattern such as a bolt pattern
center of a radius
center of a slot length or width
In all of these cases if direct tolerancing is used, confusion ma% result. !here is no reason to have this confusion except
to fall bac on =that>s the wa% we>ve alwa%s done it=. If %ou alwa%s do what %ou>ve alwa%s done, %ou>ll alwa%s get what
%ou>ve alwa%s got. In other words, %ou will never improve. !he drawing below shows the common correct application of
dimensions to center lines. In ever% case the dimensions to and from the center lines are basic.
Circularit!
(i1e , Controlling Form
Another Circularit Tip

)up3 Circularit! Error Can &e as 0arge as the Total (i1e Tolerance.
8ere is a 9uestion that comes up from time to time. =8ow can the circularit% error be as large as
the total si&e tolerance since circularit% error is a radial tolerance?= !he reason is that the @MC
and MMC si&e limits of a feature of si&e are not coaxial values. !he total si&e tolerance on the
inside diameter of 0(.:. is (.(0mm. !he figure at the bottom illustrates an extreme case where
the hole would Aust accept a 0(.0B (MMC) full form chec while at some cross sections the @MC
si&e of 0(.:, was not violated. !herefore, si&e is verified. !he circularit% error of this particular
feature is (.(0, the total si&e tolerance, since the circularit% tolerance &one is comprised of two
concentric circles. !he si&e of the circles defining the circularit% error could be 0(.0B and 0(.B as
illustrated. If this isn>t desireable, a circularit% tolerance with a value less than (.0 Could be added
to refine the circularit% control provided b% the limits of si&e. In this case, a (.()' circularit%
tolerance was added.
Concentricit! and Position , Coaxialit! Controls
Another Coa!ialit Control Tip
There are Five 8a!s to Control Coaxialit! of Features. The! are:
$ith so man% options it is no wonder that man% people find the differences hard to understand. "ach control has its
place, although some are more difficult and time consuming than others. B% comparing sample inspection methods, the
differences ma% become clearer. !his month>s tip will illustrate the differences between position and concentricit%. Cext
month>s tip will continue with the runout tolerances and profile.
Dosition ma% use the MMC and @MC modifiers on the tolerance and #atum references. !he illustration below shows a
possible gage (ignoring gage tolerance and wear allowance) that could be used when MMC is the modifier. If the si&es
are within the si&e tolerances but the part doesn>t fit the gage, the features position (coaxialit%) is out of spec. !his t%pe
of control wors well when the main concern is the assembl% of this part to another. !he gage ma% usuall% be thought of
as representing the worst mating part.
According to the A3M" 6.,.'M+.--, 3tandard, when no modifiers are present, the implied condition is regardless of
feature si&e as shown below. In this case, the axis of the datum feature and the feature being controlled must be
determined to find the error in coaxialit%. Although this control ma% be applied to bearings and d%namic balance
applications, the Aob can usuall% be accomplished at a lower overall cost b% using one of the runout controls.
All of the coaxialit% controls are intended to control concentric features. 7or that reason, man% designers and engineers
choose concentricit%. Enfortunatel%, concentricit% ignores the si&e, roundness and c%lindricit% of the feature. It re9uires
that the inspector derive a median line (see the Fune .--: !ip+of+the+Month). In a situation where %ou don>t care about
the si&e, roundness or c%lindricit% of the feature, concentricit% ma% be specified. In thirt% plus %ears of reviewing
mechanical designs, I have never found a design where this is trul% the case. !he closest application, perhaps, is when
d%namic balance is needed. In such a case, measuring a part staticall% does not assure d%namic balance if the material
is not homogeneous. If d%namic balance is re9uired, a d%namic balancing note is probabl% in order rather than
concentricit%. 7or that reason, I often use the line, =$hen in #oubt, Ese 4unout.=

Composite Tolerancing
Another Composite Tolerancin" Tip
$hen datum references are repeated in the second segment of a composite or single segment control, the
meaning is different.
!he tolerance in the upper segment of a composite tolerance is located b% all applicable basic dimensions. Gn
the drawing above, the red tolerance is located b% the red dimensions.
!he lower segment of a composite tolerance does not use the basic dimensions which originate at the datums.
Gnl% the basic dimensions within the pattern are applicable. If a datum is repeated, it indicates that the
orientation of the pattern must be held to the tighter tolerance. In this case, the perpendicularit% to datum A must
be within (.) and the pattern of two holes ma% not tilt more than (.) relative to datum B.
6ou will now notice that there are two position s%mbols being used. !his callout is not composite. It is called two
single segments. !he upper segment has the same meaning as the upper segment of the composite callout
shown earlier. !he lower segment, however, improves the location as well as the orientationof the feature(s)
relative to the datums referenced in the second sement. Cotice that the .-mm dimension from datum B is
shown in red. !he pattern must be positioned at the .-mm dimension from B within (.) total even though the
pattern ma% be out of position as much as (.0 total relative to datum C.
Cp7
Another Cp# Tip

Calculating Cp7 for 'eometric Tolerances at MMC or 0MC
!he 9uestion of calculating Cp on geometric tolerances modified at MMC and @MC continues to come up.
@ast month>s !ip showed how to inspect features of si&e where the geometric tolerance is modified at MMC
or @MC without having to find the axis or center plane. !he geometric tolerance in these cases is a single
limit control. Gnce that is reali&ed, it is eas% to calculate a Cp index on such geometric tolerances. In this
case, polar measurements are made to assure that the surface of the hole does not violate the inner
boundar%. !he inner boundar% is )( + (.. + (.0 / H.-.*. !he surface of the hole ma% not be closer to the
basic location than 4-.0'mm which is the radius of the inner boundar%. 4ecording the distribution of the
closest point relative to the inner boundar% of corresponding holes on multiple parts, provides the data
necessar% to calculate Cp of the process for the position tolerance of each hole. An% I.' calculator with
statistical functions ma% be used to calculate the average of the data, a standard deviation and the resulting
Cp.
C!lindricit! 9nspection
8hen 9nspecting C!lindricit!3 There 9s o #atum
C%lindricit% is one of the more challenging tolerances to inspect. It re9uires isolating the feature from the rest of
the part since there can never be a datum referenced with c%lindricit%. In this example, the c%lindricit% has been
applied to ' features. C%lindricit% is an individual control. !herefore, each feature is inspected independent of the
others. An electronic probe gathers man% points on the surface. A computer then evaluates the points to fit them
between two concentric c%linders that ma% not have a radial separation greater than (.(). !he actual si&e,
location and orientation of the feature are ignored. C%lindricit% is a composite of circularit%, surface straightness
and taper.
#atum $eference Frame
Another Datum $e%erence &rame Tip
#atums Are Mutuall! Perpendicular , Even 8hen the #atum Features Are ot.
It isn>t alwa%s possible or practical to select datum features that are mutuall% perpendicular to one
another when establishing a datum reference frame. Cotice that datum feature C is not nominall%
perpendicular to datum feature B. !he datum feature simulator for C would be made at *'J to the
datum feature simulator for B (shown here in red). !he actual datum planes (shown in blue), which
comprise the datum reference framewor, would however be mutuall% perpendicular to one another
as is illustrated in the last figure. !he deviation of the hole from the ''mm BA3IC location would be
measured from the third datum plane+not from the sharp point on the actual part.
#atum (hift
Another Datum Shi%t Tip
#atum Features can 'ive )ou a (hift , not a &onus.
$hen datum features are modified at MMC or @MC, the datum feature and the features being controlled ma% be able to shift relative to the datum axis or centerplane. Gn the
drawing shown below, datum feature B, in the position callout of the four hole pattern, is modified at MMC. !his means the datum simulator for datum feature B would have
a theoretical design si&e of .-.0.
3ince the actual datum feature could be produced as large as )(.), the datum feature could shift as much as
(.0 total. !his means the four+hole pattern ma% shift out of position in one direction while the datum feature
shifts in the other direction as shown. A ver% common error made at inspection is to use this allowed datum
shift as a bonus on the features being controlled. !his would allow the features to be out of position to each
other more as the datum feature departs from MMC. !his practice does not agree with the drawing or a hard
gage, should one be produced.
#atum , Patterns of Features

Patterns of Features "sed to Establish #atums Are Tric7!
The )*+.: standard tells us we can use a pattern of features to establish a datum or datums 4section +.:.;5. "nfortunatel!3 the
explanation that follows in the (tandard onl! covers a pattern of holes at MMC. There is no support for using a pattern of features of
si1e $F(. 8here two parallel holes are used as a pattern to help establish a datum reference frame3 it is prett! certain that two datum
planes will result , one plane through the two holes and a second plane midwa! between the holes. <nce more than two holes are
used as a pattern to establish a datum3 the resulting datum reference frame is unclear. This month=s Tip is to avoid using a pattern of
more than two holes to establish datums when referenced $F(.
Gn this drawing the pattern of three holes is identified as the datum
feature(s). !here is no support for how to establish this datum reference
frame 4egardless of 7eature 3i&e when inspecting the profile and other
position controls.
!he next two drawings illustrate wa%s to use two of the holes to help
clearl% establish a datum reference frame.
Primar! #atum (urface Control
Another &latness Tip
$hen a flat surface is used as a primar% datum, three considerations should be made to assure reproducible
measurement and a functional part.
.. Consider a flatness control on the datum feature. !he primar% datum is often a mounting surface that ma%
be clamped during assembl%. Clamping ma% cause strain in the part. 3train causes stress and other critical
features to var% in position or orientation. !he flatter the datum feature, the less distortion of the part.

9(PECT T>E PA$T 89T> #AT"M
FEAT"$E A
M<"TE# A'A9(T A F0AT ("$FACE
"(9'
-,M?@* &<0T( T<$A"E# T< B,*: ,m
). If the part is prone to distort and a tight flatness control is impractical, consider a constraint note. A
constraint note is usuall% intended to describe the condition of the part after assembl%.
*. If a tight flatness control or constraint note does not fit %our situation, datum targets should probabl% be
specified. 3ince three points determine a plane, three targets should be used on a primar% datum feature that is
establishing a plane. If four or more targets are used, it is advisable to also include a constraint note.
#atum , $eferencing 9tself

9t 9s $eferencing the #atum , ot 9tself
I often get the 9uestion, =8ow can a feature reference itself in a feature control frame?= !he
answer is, =It can>t.= #atum features referenced in a feature control frame are establishing a datum
reference framewor which serves as the origin of measurement. !he shaft shown has two total
runout controls that appear to be referencing themselves. !he total runout tolerance applies to the
surface of the features. !he datum is an axis which is established using the two datum features A
and B.
#atum , $epetitive Patterns

Ceep 9dentification of #atum Features (imple.
7re9uentl% parts have repetitive patterns of features. Gften it is desirable to use one of the features as a datum feature and
control the other features in the pattern relative to that datum. Gn the part shown the 0K HB holes are located from datums A, B
and C. A ver% cumbersome wa% to locate the )K H, patterns of holes relative to the HB holes individuall%, would be to assign a
different letter to each of the HB holes exhausting six letters of the alphabet. !his awward method is shown in the first drawing.
A more efficient approach is illustrated on the second drawing. !he HB holes are identified as datum feature # IC#ILI#EA@@6.
"ach pair of H, holes has a position callout relative to A, #, B where # is the large hole nearest each pair. !here are actuall% 0
datum # features. B% using the IC#ILI#EA@@6 note, the drawing is greatl% simplified.
#atum (election , &ased on Function

(elect )our #atums &ased "pon Function , ot Production
Gften production complains about the datum features selected. In general, datum features should be selected based
on function+not on how the part will be produced. !wo datum reference frames are used on the part shown below. !he
datum reference frame established b% datum features A, B and C was selected because these are the features that
locate and orient the bracet. 7rom this datum reference frame the features used to locate and orient the module are
located. !he auxiliar% datum reference frame #, ", A could then be established to position the four mounting holes.
!his approach to selecting datum features is functional and will usuall% result in the least variation in the assembl% at
the lowest cost. It is then up to production to use their sill, nowledge and expertise to produce an in+spec part.
Besides, most Droduction fols will tell %ou the% reall% don>t want the #esign fols telling them how to mae parts.
#atum Features Must &e of (ufficient (i1e
Another Datum Selection Su%%icient in Si'e
#atum selection is the most significant aspect of appl%ing 1#2!. In section ,.* of the 3tandard it
sa%s that a datum feature should be of sufficient si&e to permit its use. $ell, the maAor use of
datum features is to arrest or tae awa% the degrees of freedom or movement of a part so that it
can be reproducibl% inspected.
Gn this drawing the hole in the middle of the part is serving as the primar% datum feature. As a
c%lindrical feature, the drawing is asing this feature to tae awa% , degrees of freedom.
4idiculousM !his hole isnNt long enough to reproducibl% arrest , degrees of freedom. If %ou do this
on a drawing, please do not tell an%one %ou have been reading m% !ips.
!he second drawing has the large flat surface (which is probabl% the mounting surface) as datum
feature A. !his feature can arrest * degrees of freedom reproducibl%. !he center hole is the
secondar% datum feature. Cow it onl% has to tae awa% ) degrees of freedom which it can
reproducibl% do.
!he datum features %ou pic have to be useable. B% selecting practical datum features, %our
designs will perform better and %ou can let ever%one now that %ou are a faithful reader of m%
!ips.


Pattern , (imultaneous $eDuirements
Another Simultaneous $e(uirements Tip
9f )ou #on=t 8ant a Pattern3 )ou >ave to (a! (o%
Because of the simultaneous re9uirements rule stated in section ,.'..) in the 6.,.' 3tandard, if features are
controlled relative to the same datum reference frame, the features are considered to be a pattern. Both
drawings shown below indicate that the two e%seats are a pattern and must be located as shown.
!he e%seats are considered to be a pattern. !heir orientation is
controlled b% the position callout(s)


In order to unloc the orientation of the two e%seats to one another, a
note indicating separate re9uirements must be stated.
#atum , (!mbol Placement
Another Sm)ol *lacement Tip
8atch 8here )ou Put That Triangle%
Dlacement of the new datum feature s%mbol (triangle) can be critical. In the first three views below the
datum feature s%mbol is associated with the si&e dimension of a feature of si&e. !he% indicate that a
datum axis should be established using the feature indicated.
In the view below, the datum ma% be interpreted as a line l%ing in a plane tangent to the feature
indicated. If line contact is desired a datum target line should be indicated.
#irectl! Toleranced #imensions
Profile , 9nspection
Another Directl Tolerance+ Dimensions Tip
Toleranced #imensions (till >ave Their place.
"ven though basic dimensions are being used toda% far more than in the past, there are still man% appropriate
applications for toleranced dimensions. Care should be taen to onl% use toleranced dimensions in situations where
the% ma% be verified, reproducibl%, without the possibilit% of misunderstanding. !his generall% limits their use to<
!angent radii
3ome steps or depths
$all and material thicness
7eatures of si&e that contain opposed points

Ese of the dimension origin s%mbol clarifies the meaning
of the toleranced dimension.
Exact Constraint
Avoid <ver Constraint in )our #esigns.
Gften designers use self+aligning features in assemblies. 7or the cover shown below on the left,
two countersun holes are being used for location. Dlacing the part on datum feature A will arrest
* of the part>s 0 degrees of freedom. !ightening a flathead screw through one of the countersun
holes will arrest two more degrees of freedom. !ightening a second flathead screw would cause
over constraint and ver% often this second screwhead will not seat properl%. !o avoid this situation,
a countersun slot ma% be used in place of the second hole as is shown in the figure on the right.
!he countersin around the hole ma% then be used as a secondar% datum feature. !he
countersin on the width of the slot ma% serve as the tertiar% datum feature to provide exact
constraint during inspection and in the assembl%.
Fixed and Floating Fasteners
The Fixed and Floating Fastener Formulas are ot Eust for Fasteners.
Cearl% all assemblies fall into the categories of either fixed or floating features. A floating feature assembl% might be
a shaft floating within two bushings. $here a part floats between multiple features, the floating fastener formula will
usuall% appl%. !he A3M" 6.,.'M+.--, standard gives the floating fastener formula as 8 / 7 ; ! where<
8 is the clearance hole>s maximum material condition (smallest limit of si&e)
7 is the fastener>s maximum material condition (largest limit of si&e)
! is the positional tolerance
7or floating t%pe assemblies, Aust substitute the internal feature>s smallest si&e (MMC) for 8, the external feature>s
largest si&e (MMC) for 7 and the allowable positional tolerance (!) ma% be calculated.
$hen a screw or dowel pin is assembled, it becomes =fixed= to the mating part. !herefore, the fixed fastener
formulas wor whenever features are =fixed= on mating parts. !hese assemblies include bosses fitting into holes,
tabs fitting into slots, etc. !he A3M" 6.,.'M+.--, standard gives the floating fastener formula as 8 / 7 ; !. ; !)
where<
8 is the clearance hole>s maximum material condition (smallest limit of si&e)
7 is the fastener>s maximum material condition (largest limit of si&e)
!. and !) are the positional tolerances for the holes and fasteners.
7or fixed t%pe assemblies, Aust substitute the internal feature>s smallest si&e (MMC) for 8, the external feature>s
largest si&e (MMC) for 7 and the allowable positional tolerance (!. ; !)) ma% be calculated. !. and !) need not be
e9ual and ma% be distributed to reflect process capabilit% of the processes being used. 7or the assembl% shown
below<
!he two tabs fitting into the two slots is a fixed fastener assembl%. MMC of the two tabs is .(... MMC of the two slots
is also .(... !herefore 8 / 7 ; !. ; !) or .(.. / .(.. ; ( ; (.
Flatness
(traightness
(i1e
Consider (traightness on )our Parts Made from Flat (toc7.
Man% parts are made from stoc which is left in the as furnished condition. 4ule 5. does not appl%
to the thicness of these parts. !herefore, the form variation of the part is whatever the mill
supplies. Man% designers will control the form of such parts with flatness. 7latness is often a ver%
time consuming inspection. $hen straightness at MMC is applied to the material thicness, a
simple envelope gage at the virtual condition ma% be used to limit form variation. In this example,
the envelope gage used to verif% straightness would measure ..()'. !he actual local si&e of this
part must be checed with a micrometer t%pe measurement to assure that, at an% cross section,
the thicness is between .--' and ..(('.
Free (tate
Ese the new free state s%mbol for %our plastic, sheet metal and other flexible parts. In section .., 7EC#AM"C!A@
4E@"3 of 6.,.', paragraph (l) tells us that =All dimensions and tolerances appl% in a free state condition. !his principle
does not appl% to nonrigid parts as defined in paras. ).:...*(b) and 0.B.= !he default, then, is that all parts are
considered rigid unless there is some indication that the part ma% be flexed or distorted for inspection. A wa% to indicate
that a part is not rigid is to add a constraint note. If some tolerances are to be checed with the part not constrained, the
new free state s%mbol ma% be added to specific tolerances. Man% useable sheet metal and plastic parts will pass
inspection and be functional if clamped to a condition that simulates the assembled or functional state. "suall! when a
constraint note is applied to a drawing3 at least one tolerance should be inspected in the free state to limit the
free state variation. In this example the entire part is inspected with the part constrained except for the flatness spec
which would be checed unrestrained in the free state.
&oundaries
Thin7 &oundar!3 not &onus
!he t%pical approach to inspecting a feature such as the
n..((( hole is to measure the si&e, calculate a bonus and
then tr% to locate the axis of the hole to see if it is within the
allowable position tolerance. Another approach is to inspect
the si&e and mae measurements to see if the hole violates
the inner boundar% (virtual condition). If a hard gage were
produced to inspect the position of this hole, it would have a
pin n.-:( (;gage tolerance and wear allowance) located at the
basic location from the datums.
!his hard gage ma% be simulated using a height gage in an
open setup or CMM probe. B% adding and subtracting half the
virtual condition from the basic dimensions, the distance from
the datums that ma% not be violated ma% be calculated. !hese
distances ma% be easil% inspected.
Flatness
8hen 9nspecting Flatness3 There 9s o #atum.
7latness is another geometric tolerance that is challenging to inspect. It re9uires isolating the feature
from the rest of the part since there can never be a datum referenced with flatness. In this example,
the flatness has been applied to datum feature A. !hree possible inspection methods are illustrated. In
all cases, the considered feature is isolated from the rest of the part and aligned relative to the
indicator. In the first case, the part is leveled on the surface plate. In the second illustration, the surface
is leveled b% placing it on three e9ual height gage blocs. !he indicator is then moved across the
surface. In the third illustration, the CMM will mathematicall% =level= the points of the surface contacted
b% the probe. In all cases the 7IM (7ull Indicator Movement) ma% not exceed (.)mm.
Problem: If the surface is convex, the part will roc
maing it difficult to determine the minimum
indicator reading over the entire surface.
Problem: Ideall%, the gage blocs should be
placed under the high points on the surface.
Gtherwise, the indicator movement ma% not be
the lowest possible.
9n Examples * and - the problem ma! cause an acceptable surface to be reFected.
Gf course, a CMM will automaticall% align the
points to evaluate the flatness error.
Problem: Gften insufficient points are taen to
evaluate the flatness error. As a result, an out of
spec surface ma% be accepted.
&ottom line3 inspecting flatness reDuires
time and patience.
9nspection
(i1e , 9nspection
Profile , 0ocation of Planes

$emember3 9nspection is 8or7 Gersus $is7%
"ver%where I go I find people who thin that the part must be inspected better when 1#2! is used. Although we
tal about .((O inspection, it never reall% happens. !his example shows two wa%s to tolerance the thicness of a
bloc. 8ow can an%one declare that because a profile of a surface tolerance is used, the part must be inspected
more closel%? !he ; tolerance re9uires maing certain that no two point measurements are less than .--(. 7or
the profile of a surface, all points are to be within .(.( of the BA3IC goal of ..(((. In both cases it would not be
practical to contact all points on the surface. !herefore, enough points are checed to be confident the parts are
good. !he more points checed, the greater the confidence. In other words, it>s wor versus ris, regardless of
how the re9uirement is specified.

9nspection , >ard 'age (imulation
Another ,ar+ -a"e Simulation Tip
)ou #on=t Alwa!s eed a >ard 'age.
I fre9uentl% hear people complain that if MMC is used, the% will have to build a dedicated gage for inspection.
Gthers believe MMC is onl% appropriate when dealing with high volumes of parts. Both of these assumptions are
false. $hen MMC and @MC are applied to a geometric tolerance, the control becomes a single limit that ma% not be
violated. Gn the part below, the straightness at MMC control re9uires that the part not violate a boundar% e9ual to
the MMC si&e of ),.Bmm plus the straightness of )mm which establishes a boundar% of )0.Bmm.
!o verif% this, a simple built+up gage ma% be used as a goPno+go inspection. $here variable data is desired, an
indicator ma% be &eroed out at )0.Bmm. All readings, as the surface is trammed (probed), must be below the &ero
point. In addition, of course, the si&e must be verified with local measurements with a micrometer+t%pe device.
Cote< !he stoc si&e and material usuall% are stated in a material specification note. 7or that reason the si&e
appears here in parenthesis as a reference.

9nspection , ormal to the (urface

Measure ormal to the (urface.
$hen measuring nearl% ever% geometric control, the indicator or CMM probe should move normal (at -(J) to the surface.
!he exception is circularit% when applied to a conical surface. Circularity requires that the movement be in a plane normal
to the axis rather than the surface. If the movement for other geometric controls is not normal to the surface, a sine or
cosine correction should be made. 7or the profile of a surface control which is applicable to the )K 4B corners shown on
the drawing, once the datums are established, the part ma% be rotated about the basic location of the radii while the
indicator remains normal to the surface.

0ocation Tolerances
The Most 9mportant Control in '#HT is 0ocation3 0ocation3 0ocation.
Gnce datums have been established, other features must be located with respect to the datums. Gften those
appl%ing 1#2! will use orientation (b,f,a) and form (c,u,e2g) controls and forget to control the location.
7eatures are located geometricall% using profile of a surface (d), position(A) and, as last month>s !ip
explained, sometimes runout (h or t). !hese locating controls automaticall% provide orientation and, with the
exception of position, form control. 7orm is provided b% the si&e dimension when using position. !he
orientation and form controls should onl% be pulled out of the toolbox to be used as refinements. !he
geometric controls shown in blue come for =free= with the general Drofile of a 3urface.



MMC and 0MC
Another MMC an+ .MC Tip
Tolerances Modified at 0MC and MMC are a (ingle 0imit.
A geometric tolerance modified at MMC or @MC becomes a single limit control. !his single limit is nown as the virtual
condition. 7eatures toleranced in this manner need onl% be inspected for si&e and to assure that none of the surface of the
feature has violated the virtual condition. !his is important to reali&e when calculating Cp for a manufacturing process. !he
boundar% ma% be considered a single limit control. B% probing the feature and recording the point on the feature that is
nearest the boundar% being controlled, a Cp value ma% be determined.
7or this example, once si&e has
been checed, the hole ma% be
probed to
determine how close the surface
is to the virtual condition of
'-.'mm. B% recording these
values over time, an average and
standard deviation of the values
ma% be determined. Calculating
the difference between the
average of these values and the virtual condition
boundar%, then dividing that difference b% * standard
deviations will %ield a Cp.

otes on #rawings
)ou (till eed a Few otes.
1#2! has greatl% reduce the notes, in "nglish, that are needed on the field of the drawing. !here
are still, however a few notes that ma% be re9uired to suplement 1#2!. I have often been ased to
gather all of the notes found in A3M" 6.,.'M+.--, to provide a 9uic reference. !he following is a
table of the notes found in the 3tandard along with a brief explaination of their purpose and a
reference to where %ou can learn more about their use. 6ou might want to eep this tip close b% as
a hand% reference.
Purpose of note: (ample notes:
A(ME )*+.:M,*BB+
$eference:
&or parts that are plate+ or
coate+/ it is re(uire+ that a
note appears on the +ra0in"
statin" 0hen the
+imensional limits appl.
"xamples< #IM"C3IGCA@
@IMI!3 ADD@6 A7!"4
D@A!IC1
#IM"C3IGCA@ @IMI!3 ADD@6
B"7G4" D@A!IC1
).,..
To o1erri+e $ule 21
D"47"C! 7G4M A! MMC
CG! 4"Q#
).:.)
To esta)lish a )oun+ar o%
per%ect %orm at MMC to
control the relationship
)et0een %eatures.
D"47"C! G4I"C!A!IGC (or
CGAKIA@I!6 or @GCA!IGC
G7 36MM"!4ICA@
7"A!E4"3) A! MMC
4"QEI4"# 7G4 4"@A!"#
7"A!E4"3
).:.*
To +esi"nate the speci%ic
%eature o% a threa+/ "ear/
spline/ etc. 0hich a
MAFG4 #IA
DI!C8 #IA
MICG4 #IA
)..(
"eometric tolerance applies.
Sample note that must
appear on a +ra0in" 0hen
the sm)ol is applie+.
7"A!E4"3 I#"C!I7I"# A3
3!A!I3!ICA@@6
!G@"4ACC"# 38A@@ B"
D4G#EC"# $I!8
3!A!I3!ICA@ D4GC"33
CGC!4G@3, (G4 !G !8"
MG4" 4"3!4IC!IL"
A4I!8"M"!IC @IMI!3)
)..0.)
3hen it is necessar to
relate linear an+ an"ular
+imensions to a +atum
re%erence %rame0or#. This is
not a su)stitute %or +atum
re%erences in a %eature
control %rame.
EC@"33 G!8"4$I3"
3D"CI7I"#, #IM"C3IGC3
A4" 4"@A!"# !G #A!EM A
(D4IMA46), #A!EM B
(3"CGC#A46), AC# #A!EM
C (!"4!IA46)
,.,
To o1erri+e simultaneous
re(uirements.
3"D 4"Q! ,.'..), '.*.0.)
To o1erri+e a separate
re(uirements note or to
in1o#e simultaneous
re(uirements %or the lo0er
se"ment o% multiple
composite %eature control
%rames.
3IM 4"Q! ,.'..)..
To a1oi+ ha1in" to appl the
)asic +imension sm)ol to
each )asic +imension.
#IM"C3IGC3 @GCA!IC1
!4E" DG3I!IGC A4" BA3IC
'.)....
*lace+ )eneath a %eature
control %rame to in1o#e the
Boun+ar concept.
BGEC#A46 '..(.., 0.'.'..
To in+icate that a %eature
control %rame applies to
multiple holes sho0n on a
common a!is.
!$G CGAKIA@ 8G@"3 '......0
To a1oi+ ha1in" to appl the
)asic +imension sm)ol to
each )asic +imension.
EC!G@"4ACC"#
#IM"C3IGC3 A4" BA3IC.
7ig. 0+.)
*lace+ )elo0 a %eature
control %rame to in+icate that
the pro%ile o% a sur%ace
control applies to the entire
part.
A@@ GL"4 0.'.)
To control onl an in+i1i+ual
line element o% a sur%ace.
"AC8 "@"M"C! or "AC8
4A#IA@ "@"M"C!
0.0....
4se+ on repetiti1e patterns
to re+uce the num)er o%
+i%%erent +atums i+enti%ie+.
IC#ILI#EA@@6 '.:
To speci% 0here a tolerance
applies. This permits closer
control o% a %eature at one
sur%ace.
A! 3E47 # '.B
<rigin (!mbol
#on=t Forget the <rigin (!mbol.
Gn simple parts the dimension origin s%mbol ma% be used ver% effectivel%. !his bracet mounts on the shorter
surface as shown. A lamp is attached to the longer surface. It is desirable to inspect the *Bmm dimension b%
mounting the part on the functional shorter surface. B% using the dimension origin s%mbol, it is re9uired that this
dimension be inspected b% mounting the bracet on the shorter surface. !he same control could be accomplished
b% calling out the shorter surface as a datum feature and appl%ing a profile of a surface control on the longer
surface.

Parallelism
<rientation Tolerances #o ot 0ocate.
I eep running into people who are convinced that parallelism locates surfaces. ot True% Gn this
drawing there is a ))mm basic dimension that locates the considered surface relative to datum A.
!he parallelism tolerance has nothing to do with this dimension. All three illustrations show a
parallelism tolerance &one. !he &one is a pair of parallel planes (.)mm apart which remain parallel
to the datum simulator (the surface plate) while floating to =find= the surface. !hat>s it. !he profile
of a surface tolerance is what controls the location of the surface. !he (.' tolerance is centered on
the basic location of ))mm. !he parallelism tolerance floats inside the profile tolerance.
Inspection must =find= the surface to
determine if the surface is parallel to the
datum simulator regardless of where the
surface is. Gnl% the profile of a surface
tolerance uses the ))mm basic dimension
which provides location. $ithout the profile
tolerance the drawing would be incomplete.

You might also like